As a result of improved services and policy changes, persons with serious mental illness spend more of their lives in the community and are capable and desirous of fulfilling normalized social roles. Women with SMI are sexually active and have normal fertility and childbirth rates. However, mental health services for SMI have all but ignored parenting issues. While the effects on children of having a mentally ill mother are well-documented, research on the mothers themselves 's lacking: samples are non representative, clinical characteristics inadequately measured, the effects of culture and context on outcomes not addressed, the mothers perspective not considered, nor have changes in mothers' parenting or functioning been analyzed. The currently funded RO1 study addresses all but the longitudinal gaps in this literature. A large, multicultural sample of SMI women (N=380) has been recruited from CMHC's and in patient psychiatric units in the Detroit metropolitan area. Extensive collaboration with agencies has ensured a representative sample. Furthermore, we have established phone follow-up tracking procedures and succeeded in locating 92% of participants due for one or more 3-month recontacts. However, to more fully inform services practices, longitudinal research is needed so that variables which can only be studied over time are included (i.e., persistent economic hardship, chronic stress, changes due to children's developmental stage) and causal directions established. A conceptual model is proposed for parenting as a mediator of the relationships between stress/resources and maternal and child outcomes. Additionally, the model proposes that mh services, especially if they are parenting- Oriented, may moderate the relationships set out in the model. The current application proposes to utilize the participant sample from our original grant to carry out a 3-year longitudinal study, and reinterview P's 18 and 36 months later. Since most women's children are school-age, this time frame will encompass significant transitional periods. It should also include changes in service receipt. This should increase the likelihood of changes being experienced in the mothers' parenting and overall community adaptation. Using this longitudinal data, we propose to: a) Examine the direct impact of changes in stressors on maternal and child outcomes, controlling for the effects of personal characteristics (demographics, psychiatric history) and prior functioning; b) Examine the extent to which this relationship is moderated by mental health service provision; c) Examine the extent to which parenting attitudes and experiences mediate the impact of stressors on maternal and child outcomes; d) Analyze the extent to which mental health services (overall and parenting-specific) can moderate the relationships between stress and parenting and between parenting and maternal/child outcomes.